mxnet 1.9 build R package in windows

[YHChiuNDMC]

Hi,
I just start learning deep learning and get stocked in the last step of building R package for mxnet in windows.

Here is my working flow

• dependencies: use visual studio 2019 instead of 2022 (or CUDA cannot connect with MSVS)
• install CUDA, CuDNN, choco, OpenBLAS, mingw32-make,
• in CMD: choco install python git 7zip cmake ninja opencv pkgconfiglite
• Add on systemic Path environment variable: C:\Program Files\CMake\bin C:\MinGW\bin C:\R\R-4.1.2\bin C:\Program Files (x86)\Microsoft Visual Studio\2019\Community\VC\Tools\MSVC\14.29.30133\bin\Hostx64\x64
• open a file for mxnet building (C:\mxnet1.9)
• in CMD set to the mxnet file and download: git clone --recursive https://github.com/apache/incubator-mxnet.git -b 1.9.0
• creat a "build" file in C:\mxnet1.9\incubator-mxnet\
• in CMD set to the "build" file and make: cmake -G "Visual Studio 16 2019" -T cuda=11.5,host=x64 -DUSE_CUDA=1 -DUSE_CUDNN=1 -DUSE_CPP_PACKAGE=1 -DUSE_NVRTC=1 -DUSE_OPENCV=1 -DUSE_OPENMP=1 -DUSE_BLAS=open -DUSE_LAPACK=1 -DUSE_DIST_KVSTORE=0 -DCUDA_ARCH_LIST=Common -DCUDA_TOOLSET=11.5 -DCMAKE_CUDA_ARCHITECTURES=61 -DOpenBLAS_LIB=C:\utils\OpenBLAS\lib\libopenblas.dll.a -DOpenBLAS_INCLUDE_DIR=C:\utils\OpenBLAS\include -DCUDNN_INCLUDE=C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v11.5\include -DCUDNN_LIBRARY=C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v11.5\lib\x64\cudnn.lib -DOpenCV_DIR=C:/tools/opencv/build/x64/vc15/lib -DOpenCV_CONFIG_PATH=C:\tools\opencv\build\OpenCVConfig.cmake -DCMAKE_CUDA_COMPILER="C:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v11.5/bin/nvcc.exe" "C:\mxnet1.9\incubator-mxnet"
• open the "mxnet.sln" in visual studio
• Go to mxnet_61 project properties --> CUDA C/C++ --> additional include directories
  Add these as additional include directories :
  C:\Program Files (x86)\Windows Kits\10\Include\10.0.19041.0\winrt
  C:\Program Files (x86)\Windows Kits\10\Include\10.0.19041.0\um
  C:\Program Files (x86)\Windows Kits\10\Include\10.0.19041.0\ucrt
  C:\Program Files (x86)\Windows Kits\10\Include\10.0.19041.0\shared
  C:\Program Files (x86)\Windows Kits\10\Include\10.0.19041.0\cppwinrt\winrt
  C:\Program Files (x86)\Microsoft Visual Studio\2019\Community\VC\Tools\MSVC\14.29.30133\include
• Build (Ctrl + Shift + B) => take about 4 hours
  (Then I was stocked here)
• in CMD: cd C:\mxnet1.9\incubator-mxnet
• make R-package/Makefile rpkg
  process_begin: CreateProcess(NULL, uname -s, ...) failed. process_begin: CreateProcess(NULL, uname -a, ...) failed. 系統找不到指定的路徑。 process_begin: CreateProcess(NULL, uname -s, ...) failed. process_begin: CreateProcess(NULL, cc -dumpmachine, ...) failed. process_begin: CreateProcess(NULL, cc -dumpmachine, ...) failed. Makefile:256: "USE_LAPACK disabled because libraries were not found" Makefile:384: WARNING: Significant performance increases can be achieved by installing and enabling gperftools or jemalloc development packages 找不到檔案 - TESTS/CPP/ 找不到檔案 - -NAME Makefile:769: warning: overriding recipe for target 'build/libcustomop_gpu_lib.so' Makefile:757: warning: ignoring old recipe for target 'build/libcustomop_gpu_lib.so' Makefile:772: warning: overriding recipe for target 'build/libsubgraph_lib.so' Makefile:760: warning: ignoring old recipe for target 'build/libsubgraph_lib.so' make: Nothing to be done for 'R-package/Makefile'. make: *** No rule to make target 'rpkg'. Stop.

I tried to performed rpkg commend by hand (It seems linux commend) but still cannot install to R.
-creat file /R-package/inst/libs
-copy "/src/io/image_recordio.h" to "/R-package/src"
-(cp -rf lib/libmxnet.so R-package/inst/libs
if [ -e "lib/libtvm_runtime.so" ]; then
cp -rf lib/libtvm_runtime.so R-package/inst/libs;
fi) => cannot find lib file
-creat file /R-package/inst/include
-copy everything in "include" file to "R-package/inst/include"

• I already installed devtools libcairo2-dev libxml2-dev roxygen2 in R.
• change filename /R-package/dummy.NAMESPACE to /R-package/NAMESPACE
• in CMD: echo "import(Rcpp)" >> R-package/NAMESPACE
• in CMD: R CMD INSTALL R-package =>failed

Can somebody help me with the make rpkg step? Many thanks!

My system info:
` sessionInfo()
R version 4.1.2 (2021-11-01)
Platform: x86_64-w64-mingw32/x64 (64-bit)
Running under: Windows 10 x64 (build 19042)

Matrix products: default

locale:
[1] LC_COLLATE=Chinese (Traditional)_Taiwan.950
[2] LC_CTYPE=Chinese (Traditional)_Taiwan.950
[3] LC_MONETARY=Chinese (Traditional)_Taiwan.950
[4] LC_NUMERIC=C
[5] LC_TIME=Chinese (Traditional)_Taiwan.950

attached base packages:
[1] stats graphics grDevices utils datasets methods base

other attached packages:
[1] devtools_2.4.3 usethis_2.1.5

loaded via a namespace (and not attached):
[1] magrittr_2.0.1 pkgload_1.2.4 R6_2.5.1 rlang_0.4.12
[5] fastmap_1.1.0 fansi_0.5.0 tools_4.1.2 pkgbuild_1.3.0
[9] sessioninfo_1.2.2 utf8_1.2.2 cli_3.1.0 withr_2.4.3
[13] ellipsis_0.3.2 remotes_2.4.2 rprojroot_2.0.2 tibble_3.1.6
[17] lifecycle_1.0.1 crayon_1.4.2 processx_3.5.2 purrr_0.3.4
[21] callr_3.7.0 vctrs_0.3.8 fs_1.5.2 ps_1.6.0
[25] testthat_3.1.1 memoise_2.0.1 glue_1.6.0 cachem_1.0.6
[29] pillar_1.6.4 compiler_4.1.2 desc_1.4.0 prettyunits_1.1.1
[33] pkgconfig_2.0.3
list.files()
[1] "3rdparty" "amalgamation" "appveyor.yml"
[4] "benchmark" "build" "cd"
[7] "ci" "cmake" "CMakeLists.txt"
[10] "CODE_OF_CONDUCT.md" "CODEOWNERS" "config"
[13] "contrib" "CONTRIBUTORS.md" "cpp-package"
[16] "DISCLAIMER" "doap.rdf" "docker"
[19] "docs" "example" "include"
[22] "julia" "KEYS" "LICENSE"
[25] "licenses" "make" "Makefile"
[28] "matlab" "mkldnn.mk" "MKLDNN_README.md"
[31] "NEWS.md" "NOTICE" "perl-package"
[34] "plugin" "pytest.ini" "python"
[37] "R-package" "rat-excludes" "README.md"
[40] "readthedocs.yml" "scala-package" "SECURITY.md"
[43] "setup-utils" "snap.python" "snapcraft.yaml"
[46] "src" "tests" "tools"
getwd()
[1] "C:/mxnet1.9/incubator-mxnet"`